Fintech company Klarna, which powers e-commerce transactions for some of the world’s most recognizable brands, including Expedia, Macy’s and Nike, is at the forefront of AI adoption. It has integrated artificial intelligence across the company, most notably with an AI chatbot that it recently said does the equivalent work of 700 customer service agents. Klarna, which employs roughly 4,000 people, recently released statistics that show how efficient and effective the tool has been, wading into the thick of sensitive and high-stakes debates about the role of generative AI in business, how humans interact with it and its implications for the future of work. CEO Sebastian Siemiatkowski explains why he is so transparent about AI’s capabilities, and what concerns him most about the new technology. This interview has been edited for length and clarity.
Bloomberg
AI’s takeover of human jobs is a touchy topic. Why did you choose to publicly share data on Klarna’s use of AI to replace customer service agents?
We worry in general about the effects this may have on society, so we decided to be upfront about the fact that it has had some amazing outcomes for customers.
It is currently doing the equivalent work of about 700 full-time [customer service] agents. Klarna does not employ customer service agents ourselves, we use some of the large customer service providers out there. They’re outsourced — they are not employees.
We made the announcement to say the consequence of us launching the technology is we need the equivalent of 700 fewer full-time agents than what we usually use on an average basis. On average, we need 3,000 agents, now we need a little more than 2,000.
We wanted to make policymakers aware that this isn’t something that’s happening in the future — it’s happening now. We think it is critical that society start thinking about this major change.
So this isn’t related to the layoffs Klarna conducted in 2022? AI didn’t take those jobs?
In 2022, as a consequence of a change in investor sentiment, we had to reduce the size of our company. We reduced our staff by around 800 people. None of [the people we laid off] were customer service agents. Now, two years later, we are looking at this interesting technology we’ve developed with ChatGPT. These are two totally separate things that have been conflated.
In the short-term, there are no layoffs or implications for employees as a result of us launching this customer service AI chatbot.
What does the chatbot do? How do you measure its success?
It handles two-thirds of our customer service chat inquiries. It’s on par with humans in terms of satisfaction and it resulted in a 25% reduction in repeat inquiries from customers.
Before we took this live, we already had a co-pilot that helped customer service agents and other employees accelerate their work. This AI actually communicates and resolves customers’ issues on its own. We think it’s important that people are still given the option to speak to a human, but we’re also seeing that people who choose to interact with the AI chatbot are very happy and find it helpful, to the point and effective. It can communicate in 35 languages, so for many immigrant and expat groups, it means a huge improvement in their experience.
In large organizations, less time is spent on what really creates value for employees, customers and shareholders. With AI, it’s the less-productive work that can be taken away. In the best of worlds, everyone who comes into the company will be able to use their creative power to create real value for customers — and not to write an impressive presentation for its CEO. That’s less of a value add; it’s specifically that kind of work I hope to shrink.
Do you worry about it making mistakes or delivering inaccurate information?
One has to remember that unfortunately, it’s not like we humans are perfect. Humans are fantastic but they also make mistakes, either because they didn’t [give a query] proper attention or get training, and it’s not always their fault. The point is, it does happen, especially when doing millions of customers’ inquiries, some aren’t as good as you’d like them to be.
The key metric we use is making sure it makes fewer mistakes, on average, than humans do. That’s the objective and it’s something we’ve been monitoring very closely. We make sure the AI’s mistakes are less severe and less common.
How has AI changed your approach to hiring, if at all?
We’ve stopped hiring in the last six months. We’re shrinking as a company, not by layoffs, but by natural attrition. Klarna tries to apply AI across all products and services and work we do. It’s having implications on how many people we need as a company. This is one time that a single product improvement led to a massive reduction in need for customer service agents.
How do your employees feel about this?
As much as it’s not easy to not talk about it, we want to be honest and transparent and tell people about it because of the implications for society. As far as our own employees go, we have been trying to communicate this internally since we stopped recruiting in October — that we believe this will allow us to do more with less.
Our ambition is to invest more per employee and to see the compensation of existing employees go up as we become a higher-revenue company.
Internally, lots of employees have been excited about AI and are applying it and Klarna, which is known as a leading AI company, which I think makes our workers more attractive to other employers.
How did you reach your estimation of AI’s $40 million in cost savings to Klarna?
We were buying customer service from suppliers to take care of inquiries. We had an amount budgeted for how much we were going to spend with those suppliers this year. Now, thanks to the technology, we will spend much less. Forty million is the delta between what we expected to spend and what we will actually spend.
What do you think the larger of implications of AI for society will be?
We are sharing this information to pose the question, to society at large, of how we are going to manage this massive change.
From what I’ve heard, it’s less about job security and more about general security, as in, we’ll need electronics IDs to know that we are in fact humans talking to one another.
Also, while new [AI-related] jobs may be created on a societal level, that’s not a solve for the individual [who is replaced by AI].
The better we as a society can support individuals who may be affected, the more positive a development it will be. The idea is not to stop it; I don’t think that’s right. The question is, ‘What support can we give people who are affected by it?’